Maternal health behavior in the DRC

UNCIEF MICS data

Fuyu Guo

October 20, 2020

MICS: Multiple Indicator Cluster Surveys

Dataset Structure

MICS is a household survey, includes two levels information:

We used household and individual women datasets.

Dataset Structure

Household dataset

Inidividual women dataset

Dataset Structure

library(dplyr)
hh <- read.csv("D:\\Tang_new pooled model\\MICS4-6\\hhmics4.csv")
head(hh[, c("HH1", "HH2")])
  HH1 HH2
1   1   1
2   1  11
3   1  21
4   1  30
5   1  40
6   1  50
dim(hh)
[1] 11490    13

wm <- read.csv("D:\\Tang_new pooled model\\MICS4-6\\wmmics4.csv")
head(wm[, c("WM1", "WM2")])
  WM1 WM2
1   1   1
2   1  11
3   1  11
4   1  21
5   1  30
6   1  30
dim(wm)
[1] 13235    27
hh <- mutate(hh, hID = paste(hh$HH1, hh$HH2, sep="_"))
wm <- mutate(wm, hID = paste(wm$WM1, wm$WM2, sep="_"))
df <- merge(hh, wm, by = "hID")
head(df[, c("hID","HH1","HH2","WM1", "WM2")])
##     hID HH1 HH2 WM1 WM2
## 1   1_1   1   1   1   1
## 2  1_11   1  11   1  11
## 3  1_11   1  11   1  11
## 4 1_118   1 118   1 118
## 5 1_127   1 127   1 127
## 6 1_127   1 127   1 127
dim(df)
## [1] 13235    41

Special variable

Below variables are not mentioned in the survey, but are calculated and provided in the dataset

Project 1 Maternal seeking behavior in MICS6

We aimed to study maternal seeking behavio (MHSB) in the DRC.
MHSB includes:

Questions about these tree questions were answered by women who gave live birth in the last two years

PNC : We need to be very careful on its definition and questions

Project 1 Maternal health seeking behavior in MICS6

We aimed to study the maternal health seeking behaivor of women in MICS6.
By conducting a k-modes cluster analysis, we categorized women into three groups:

Highly-used Group Insufficiently-used Group Seldom-used group
ANC1 (%) 95.4 100.0 37.0
ANC4 (%) 63.0 43.0 11.3
Institutional delivery (%) 92.8 100.0 14.7
Examine the cord (%) 90.5 7.0 7.7
Examine the temperature (%) 59.8 0.7 0.6
Counsel on breastfeeding (%) 69.9 2.6 4.6

Project 1 Maternal health seeking behavior in MICS6

Percentage of women in three groups in the DRC

Project 1 Maternal seeking behavior in MICS6

Geographic distribution of three groups in the DRC

Project 1 Maternal seeking behavior in MICS6

Covariates Seldom-Used Group versus (Highly-Used Group & Insufficiently-Used Group) (Insufficiently-Used Group & Seldom-Used Group) versus Highly-Used Group
Education level
Below Primary Ref Ref
Primary 1.36 (1.17-1.57) 1.36 (1.17-1.57)
Secondary or higher 2.20 (1.84-2.63) 1.57 (1.31-1.88)
Wealth Group
Pooresr Ref Ref
Poor 1.40 (1.25-1.57) 1.40 (1.25-1.57)
Median 1.46 (1.26-1.69) 1.46 (1.26-1.69)
Wealthy 2.10 (1.60-2.75) 1.43 (1.54-3.40)
Wealthiest 2.28 (1.54-3.40) 2.28 (1.54-3.40)
Note:
All other demographic variabls are controled

Project 2 Trend for MHSB in the DRC

In the second project, we studied the trend of maternal health service coverage trend from 2010 to 2018 by using MICS4 and MICS6.
MICS4 and MICS6 had similar sampling and survey structure.

mics4 <- read.csv("D:\\Tang_new pooled model\\MICS4-6\\Merged Data\\merge_mics4.csv")
mics6 <- read.csv("D:\\Tang_new pooled model\\MICS4-6\\Merged Data\\merge_mics6.csv")
table (mics4$province)

          Bandundu          Bas congo           Equateur   Kasai Occidental 
              1213                994               1106               1147 
    Kasai Oriental            Katanga           Kinshasa            Maniema 
              1114               1244               1482               1101 
         Nord Kivu Province Orientale           Sud Kivu 
              1247               1082               1121 
table (mics6$province)

      Bas Uele       Equateur   Haut Katanga    Haut Lomami      Haut Uele 
           270            288            400            445            256 
         Ituri          Kasai  Kasai Central Kasai Oriental       Kinshasa 
           291            390            370            350            255 
 Kongo Central         Kwango          Kwilu         Lomami        Lualaba 
           263            340            240            396            317 
     Maindombe        Maniema        Mongala      Nord Kivu    Nord Ubangi 
           260            324            306            339            358 
       Sankuru       Sud Kivu     Sud Ubangi     Tanganyika         Tshopo 
           383            389            381            260            352 
       Tshuapa 
           334

Project 2 Trend for MHSB in the DRC

Most questions could be found in both of two datasets, but the variable name may differ. For an example, has the women received skilled attendance at delivery by a doctor

table(mics4$MN17A)
## 
##          Manquant  Médecin 
##    12445        2      404
table(mics6$MN19A)
## 
##         A 
## 8086  471

Project 2 Trend for MHSB in the DRC

Change of adjusted coverage of ANC (left) and skilled attendance at delivery (right)

Project 2 Trend for MHSB in the DRC

change of maternal health service coverage amont different SES

Github link https://github.com/fyGuo/MICS4-6

Thanks